I’m discussing this on the forums, so I thought I would write something here too. The important thing is that what applies to this game is going to happen again and again.
Here’s a schematic approach to how to judge “No Man’s Sky” and understand the debate going on right now.
1- The “over-hype”. There’s lot of discussion about the hype of this game. Every review will mention this. People will argue endlessly about who’s responsible of this hype and whether the game is good or bad depends strictly on how close it lands to someone’s expectations. As if good or bad depends solely on expectations management.
Answer: the hype doesn’t depend on Sony, or the dev team, or the internet. The hype of this game is caused by the use of “procedural generation”. It’s this piece of tech that carries a baggage of “over-hype” and every future game that heavily relies on procedural generation to build its content will face a similar over-hype. It has embedded the myth: “with just a few rules (or a small team) I can create an almost infinite UNIVERSE that you can have fun to explore endlessly”.
The idea that you can produce a large amount of content with little effort is just plain stupid. It’s the opposite: procedural generation requires MORE work to be good or on par with handcrafted content. Dwarf Fortress is great because it’s been developed non-stop for more than 10 years. There are no shortcuts, there’s no magical formula to produce interesting content.
2- This game uses procedural generation in a very stupid way.
There’s a “good” and “bad” use of procedural generation (tech is not good or bad on its own, it’s the use that matters), and it’s also easy to analyze since it depends on a simple thing. “Good” procedural generation makes the environment dictate gameplay. If the player has a plan, or a list of activities to optimally reach a goal, then for every new game that plan might be followed closely to produce the optimal result. But if you instead “procedurally generate” the environment and make it the center of the experience then it means you force the player to observe and adapt. Not anymore you arrive with a pre-made plan, but your strategy needs to adapt and learn from what you find. Every time the context changes, so every time the experience changes too. This is also the seed for interesting “exploration”. It’s not simply about sightseeing, it’s about making gameplay be shaped by the experience. Changing radically that experience. You transform the environment because you want the environment to transform gameplay (where the best result becomes “emergent”, in those rare cases when the rules are really solid).
But instead this game uses the “bad” kind of procedural generation, which is: cosmetic variations of functionally identical elements. This game is all about stuff looking slightly different but having the exact same function. You travel to a new planet, the previous one had a bluish tinge, this one a greenish one, but what you actually do on every planet is repetitive. You shoot a different looking rock or a plant to obtain the same material. It’s as if every object in this endless world is a box containing the same content, but a different shape on the outside.
It’s the same “sin” in Oblivion: what’s the point of “exploring” and finding a dungeon hidden at the border of the map when the spawn list of what’s inside is always the same? You’re going to find in that far-away dungeon the exact same content because they share the same exact spawn list.
Since this game uses procedural generation mostly for cosmetic reasons, the result is that the gameplay feels “dull”. Not because “there’s not enough to do”, which is what has been discussed for months, but because what you do DOESN’T DEPEND ON WHAT YOU ARE GOING TO FIND. That’s the “sin” of this game, the core of its bad game design. The gameplay function is independent from the procedural generation. There’s a total disconnect between the “exploration” and function. Between what you see and what you use.
Every planet LOOKS different but PLAYS the same. This is bad game design, and a very bad use of procedural generation.
3- Deliberate scamming: https://www.youtube.com/watch?v=AE0nuW-mQ8A
It was one of the most frequent questions whether or not you could meet another player in the game. They always answered vaguely, stating it’s not what a player “should be looking for”. But the question is very precise, did you write that code or not? Can you see another player or not? Turns out you just cannot. No Man’s Sky is entirely single-player. The code is just not there. There’s some indirect information that goes back and forth, so you could see what another player named that planet he discovered, but that’s the limit. It’s not that you cannot see another player because the universe is too vast to meet, which is what they said over and over, you cannot see another player because the networking code just wasn’t written.
And that’s one sign that says that, even if the over-hype is mostly due to a misuse of tech, the devs themselves heavily exploited that over-hype to sell the idea of a game over the actual game. They chose marketing over honesty.
They have a small team, they’ll make tons of money thanks to that hype. The game is a success. But they’ll also deservedly earn a very nice amount of bad reputation because of this, and maybe in the future players will be a bit less gullible to the false magic of procedural generation sprinkled by good programmers turned into bad game designers.
“No Man’s Sky” is a cute piece of tech wrapped around very bad game design.
Disclaimer: most of this was written *before* the game’s release. The point was to determine how to judge between a good and bad use of procedural generation. My comments about the game come from what I read and saw about the game in these first few hours, so you can judge by yourself if what I wrote also holds up as a complete description of the actual game.
Why is this important, why not wait before judging? Because opinions on the internet are entirely worthless. It doesn’t matter what *I* think about the game, it doesn’t matter what *you* think either. What matters and is worth writing and reading is about motivations. It’s about the discussion on the whys and hows. So here you can see my thought process while judging the game. It doesn’t matter if I think its game design is bad, what matters is that I described what, potentially, makes its game design bad. And all that stays valid even if eventually the game turns out differently.
Disclaimer bis: The “scam” of the title refers to the fact no one is actually responsible for the “scam”, or the over-hype, or the chimera. It was all embedded in the misperception and misuse of the procedural generation tech. That the devs have very deliberately and maliciously exploited just to make more money rather than offer an honest image of the game. It’s just intellectual dishonesty, of course. It’s pretty pervasive. I remember at least one occasion commenting on building hype by exploiting false myths. It was Vanguard, or selling the vagueness of an idea so that people’s mind would fill the picture with whatever it is they love.